The Philippines Just Told Every Public School: Use AI. Here Is What DepEd Order 003 Actually Says.

The Philippines Officially Embraces AI in Public Schools: What DepEd Order 003 Really Means

When a government tells 17 million students they can use artificial intelligence in their classrooms, it's not a casual suggestion. In February 2026, the Philippines' Department of Education issued Department Order No. 003, a 49-page directive that transforms how public schools approach AI. This isn't a ban disguised as regulation. It's an explicit permission slip with guardrails^✦.

The timing matters. While some countries still debate whether ChatGPT belongs in schools, the Philippines moved decisively. DepEd Secretary Sonny Angara framed it clearly: "There's an ethical use of AI. We will not prohibit the use of AI." Instead, the order requires something more nuanced: disclosure and graduated access based on student maturity.

This is what the policy actually does, and why it signals a shift in how Southeast Asia thinks about education technology.

How the Philippines Approved AI Tools (And Why)

The order approves seven specific tools: ChatGPT, Google Gemini, Grammarly, Quillbot, Canva, Tome, and Khanmigo. These aren't random choices. Each addresses a defined educational use case, from writing assistance to presentation design to personalised tutoring via Khan Academy's AI chatbot.

But approval comes with conditions. All tools must meet "ethical, pedagogical, and human-centred standards." The framework divides AI applications into risk categories:

• Low-risk uses (grammar checking, spam filtering, chatbot tutoring) face minimal restrictions
• High-risk applications (grading, admissions decisions, scholarship allocation) require strict human oversight

This graduated approach recognises reality: AI is more trustworthy as a writing assistant than as an admissions officer.

The Core Requirements: Disclosure, Not Prohibition

Angara explained the philosophy in a way that cuts through the usual hand-wringing: "We just need the students to disclose how they use AI so that it can be integrated into their learning. It's a graduated policy; the more independent learners will be given greater latitude."

That's the operating principle. Students must be transparent about AI use in assignments. Teachers, not algorithms, decide what constitutes appropriate application. Independent learners get more flexibility. Younger or less experienced students get clearer boundaries.

The policy treats AI as a support tool, never as a replacement for teacher judgment. This distinction separates it from purely laissez-faire approaches that let the technology drive pedagogy.

What's Prohibited: Hard Lines on Risk and Ethics

The order draws clear red lines. Schools cannot use AI for:

• Biometric categorisation or facial recognition scraping
• Emotion detection systems that profile students
• Social scoring mechanisms that label learners
• Manipulative chatbots designed to target minors
• Automated decisions on grading, admissions, or scholarships without human review

These prohibitions reflect growing global concern about algorithmic bias^✦ and surveillance in education. The Philippines didn't create these rules in isolation. Similar guardrails appear in Singapore's AI governance^✦ work and emerging frameworks across Southeast Asia.

Permitted Uses: What Teachers and Students Can Actually Do

The order explicitly permits AI for:

• Creating instructional materials and lesson plans
• Analysing educational data to identify learning gaps
• Grammar checking and citation verification
• Evaluating AI tools for exam security and reliability
• Summarising research for student projects
• Generating practice questions and formative assessments

These uses reflect how teachers actually want to use AI: as a lever to reduce busywork and create space for higher-order thinking. The framework acknowledges this doesn't happen if teachers spend hours hand-grading essays or manually checking citations. Similar initiatives are underway across Asia: Microsoft has launched a programme to train two million Indian educators on AI skills, while Vietnam is running a 170-school AI education pilot in Ho Chi Minh City to prepare the next generation.

There's an ethical use of AI. We will not prohibit the use of AI.

The Broader Context: Project AGAP.AI and the Southeast Asian Momentum

DepEd Order 003 didn't emerge in a vacuum. In January 2026, the Philippine government launched Project AGAP.AI (Accelerating Governance and Adaptive Pedagogy through AI) with President Ferdinand "Bongbong" Marcos Jr. The initiative positions AI as infrastructure for educational equity, not privilege.

This reflects a regional trend. Singapore released its comprehensive AI governance framework for higher education last year. Thailand launched the Google AI Literacy Academy to upskill educators. Malaysia moved from voluntary guidelines to binding legislation on AI governance. Cambodia published its digital and AI readiness strategy. The Philippines' order fits into this broader Southeast Asian push to integrate AI thoughtfully rather than resist it reflexively.

Students need to disclose how they use AI so that it can be integrated into their learning.

How Schools Will Implement This in Practice

The real test isn't the policy's wording. It's whether teachers can actually implement it. DepEd Order 003 doesn't hand implementation to AI. It hands it to teachers, who must:

• Develop assessment rubrics that account for appropriate AI use
• Create assignment types that benefit from AI disclosure
• Maintain human judgment over all high-risk applications
• Monitor for misuse without becoming surveillance specialists
• Balance student autonomy with pedagogical responsibility

This creates work. But it also creates space for genuine pedagogical innovation rather than dystopian surveillance scenarios that sometimes dominate AI-in-education debates.